DESCRIPTION: The aim of the proposed project is to test a longitudinal social context model of the development of alcohol use during pre-adolescence and adolescence across White and African American youth. Within the theoretical model, family, peer, neighborhood, and school influences are hypothesized to be important risk factors for the development of alcohol use. Specifically, adolescent alcohol initiation and development is hypothesized to be influenced by family conflict/cohesion, parenting practices, and family substance use, as well as associations with deviant peers and peer use of substances. School and neighborhood contextual characteristics such as socio-economic status (SES), crime rates, drug problems, and social disorganization are also hypothesized to influence adolescent alcohol development. Homogeneity and heterogeneity in alcohol use and other substance use across White and African American families and the influence of contextual factors on family levels of substance use will also be examined. In addition to focusing on the risk and protective factors contributing to the development of alcohol use, the present study will also assess the dynamic relationships between alcohol, other substance use, and related problem behaviors such as academic failure, antisocial behavior, and risky sexual behavior. Statistical methods primarily designed to explore development and growth, and multilevel data, will be employed. Specifically, latent growth curve methodology, generalized linear modeling, and hierarchical or multilevel modeling techniques will be employed to determine the influence of each of the antecedent risk and protective factors on adolescents' development of alcohol use and other concomitant problem behaviors. The structural relationships among these factors and their influences on the developmental process within the context of the hypothesized and other competing developmental models will be explored and compared across African American and White adolescents and families.